How much data to optimize our queries - sql-server

Question

i have one table with this stracture and Indexs
CREATE TABLE [dbo].[Report4](
[Id] [int] IDENTITY(1,1) NOT NULL,
[Mesc] [nvarchar](50) NULL,
[Line] [nvarchar](5) NULL,
[Unit] [nvarchar](5) NULL,
[Description] [nvarchar](500) NULL,
[ST_CODE] [nvarchar](5) NULL,
[PbsNo] [nvarchar](50) NULL,
[PbsDate] [nvarchar](10) NULL,
[PbsQty] [nvarchar](10) NULL,
[PbsQtyRec] [nvarchar](10) NULL,
[QtyConsum1] [nvarchar](10) NULL,
[QtyConsum2] [nvarchar](10) NULL,
[QtyConsum3] [nvarchar](10) NULL,
[QtyConsum4] [nvarchar](10) NULL,
[QtyConsum5] [nvarchar](10) NULL,
[Type] [nvarchar](20) NULL,
[InvQty] [nvarchar](10) NULL,
[TypeRequest] [nvarchar](50) NULL,
[HeaderId] [bigint] NULL,
[LOCATION] [nvarchar](10) NULL,
CONSTRAINT [PK_Report4] PRIMARY KEY CLUSTERED
(
[Id] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [HeaderId] ON [dbo].[Report4]
(
[HeaderId] ASC
)
INCLUDE ( [Id],
[Mesc],
[Line],
[Unit],
[Description],
[ST_CODE],
[PbsNo],
[PbsDate],
[PbsQty],
[PbsQtyRec],
[QtyConsum1],
[QtyConsum2],
[QtyConsum3],
[QtyConsum4],
[QtyConsum5],
[Type],
[InvQty],
[TypeRequest]) WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [HeaderIdRAll] ON [dbo].[Report4]
(
[HeaderId] ASC
)
INCLUDE ( [Id],
[Mesc],
[Line],
[Unit],
[Description],
[ST_CODE],
[PbsNo],
[PbsDate],
[PbsQty],
[PbsQtyRec],
[QtyConsum1],
[QtyConsum2],
[QtyConsum3],
[QtyConsum4],
[QtyConsum5],
[Type],
[InvQty],
[TypeRequest]) WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [LineNOIRHeaderId] ON [dbo].[Report4]
(
[Line] ASC
)
INCLUDE ( [HeaderId])
WHERE ([line]='I')
WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [LineNoRHeaderId] ON [dbo].[Report4]
(
[Line] ASC
)
INCLUDE ( [HeaderId])
WHERE ([line]='H')
WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [LineNoRMesc] ON [dbo].[Report4]
(
[Line] ASC
)
INCLUDE ( [Mesc])
WHERE ([line]='I')
WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
CREATE NONCLUSTERED INDEX [MescRAll] ON [dbo].[Report4]
(
[Mesc] ASC
)
INCLUDE ( [Id],
[Line],
[Unit],
[Description],
[ST_CODE],
[PbsNo],
[PbsDate],
[PbsQty],
[PbsQtyRec],
[QtyConsum1],
[QtyConsum2],
[QtyConsum3],
[QtyConsum4],
[QtyConsum5],
[Type],
[InvQty],
[TypeRequest],
[HeaderId]) WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO
i want run this Query for this table
ALTER PROCEDURE [dbo].[SPSelectReport2] (#StringWhereParameter nvarchar(4000),#PageIndex int,#PageSize int )
AS
BEGIN
SET NOCOUNT ON;
-- َ Begin Of Transaction
begin tran
declare #from int=(#PageSize*(#PageIndex-1))+1
declare #to int=(#PageIndex*#PageSize)
declare #Query nvarchar(max)
set #Query=' select
distinct id, [Mesc], [Line]
,[Unit]
,[Discription]
,[InvQty]
,[LastDateNil]
,[StCode]
,[PlanCode]
,[MIN]
,[MAX]
,[LastDateConsum]
,[PbsNo]
,[PbsDate]
,[PbsQty]
,[PbsQtyRec]
,[DateDelay]
,[TypeRequest]
,[HeaderId]
,LOCATION
from (
SELECT *, ROW_NUMBER() OVER(ORDER BY Id) ROW_NUM
FROM(
((SELECT Id,[Mesc]
,[Line]
,[Unit]
,[Discription]
,[InvQty]
,[LastDateNil]
,[StCode]
,[PlanCode]
,[MIN]
,[MAX]
,[LastDateConsum]
,[PbsNo]
,[PbsDate]
,[PbsQty]
,[PbsQtyRec]
,[DateDelay]
,[TypeRequest]
,[HeaderId]
,LOCATION
FROM [MyMaterialDB].[dbo].[Report2]
WHERE headerid IN(SELECT HeaderId FROM [MyMaterialDB].[dbo].[Report2] WHERE line=''H'''+ #StringWhereParameter+'))
UNION
(
(SELECT Id,[Mesc]
,[Line]
,[Unit]
,[Discription]
,[InvQty]
,[LastDateNil]
,[StCode]
,[PlanCode]
,[MIN]
,[MAX]
,[LastDateConsum]
,[PbsNo]
,[PbsDate]
,[PbsQty]
,[PbsQtyRec]
,[DateDelay]
,[TypeRequest]
,[HeaderId]
,LOCATION
FROM [MyMaterialDB].[dbo].[Report2]
WHERE mesc IN(SELECT mesc FROM [MyMaterialDB].[dbo].[Report2] WHERE line=''I''' +#StringWhereParameter+'))
UNION
(SELECT Id, [Mesc]
,[Line]
,[Unit]
,[Discription]
,[InvQty]
,[LastDateNil]
,[StCode]
,[PlanCode]
,[MIN]
,[MAX]
,[LastDateConsum]
,[PbsNo]
,[PbsDate]
,[PbsQty]
,[PbsQtyRec]
,[DateDelay]
,[TypeRequest]
,[HeaderId]
,LOCATION
FROM [MyMaterialDB].[dbo].[Report2]
WHERE mesc IN(SELECT HeaderId FROM [MyMaterialDB].[dbo].[Report2] WHERE line=''I'''+#StringWhereParameter+')
)))) a)b where b.ROW_NUM between '+CAST(#from as varchar(10))+' and '+CAST(#to as varchar(10))
-- Order by Mesc,Line,unit
exec(#Query)
--print #Query
--
if ##error = 0
Commit Tran
Else
rollback tran
End
into this table i have more then 1000000 records and when run this sp ,It takes ten minutes or more. How do I optimize the structure or query. please help me. thanks all.

Answer 1

For starters you MUST do the following (I think only those actions can take your query execution time down to much less than a minute):
remove every in clause. select ... from ... where a in (select ...) is a) very very very slow, b)dangerous as it can run your sql server out of memory -especially for large datasets like yours-. You must change all in clauses to left outer joins, it is very easy to do this.
remove distinct clause. This forces your query to do order by (behind the schenes) and generally is very slow too. Try to rephrase your query so that records are already distinct
As long as you have distinct data you must also change union to union all. Union is same as distinct union which goes as my previous statement. If you use union all you will see great improvement to your query... as long as you can bring your data to not needed to be distinct...
If you manage to make at least 2 of the 3 previous changes, i think you will see great improvement. If you can handle all of them, you will have a very fast query.
And also, do not bother changing your sql exec into a 'precompliled' stored procedure. I don't think you will gain an significant improvement by doing this (I see you have dynamic where so this wouldn't be applicant in any case).
Please, tell me if that worked.